Last week I noted some of the problems with trying to attack inequality by taxing the wealth, or the unrealized capital gains, of billionaires: When you tax something heavily, you generally get less of it. And capital isn’t something we want less of; capital is what gets invested to make the economy more productive. A more productive economy means we all get more stuff for less work — and even if you think “more stuff” sounds dubious, surely “less work” has some appeal.

But some sort of tax hikes are definitely coming, because the U.S. fiscal gap is too large to sustain indefinitely. And many people figure the folks sitting on giant piles of wealth are the ones who can most easily afford tax hikes.

They’re right. But given the flaws of capital taxation, I’d ask: Is it their wealth we want to redistribute? Or is a better target the real goods and services they use that wealth to consume?

This is the point economist Steven Landsburg made in his 2011 essay “The Man Who Can’t Be Taxed.” Landsburg offers the example of a rich heir who spends his days fiddling with a small collection of derelict luxury cars. I certainly understand why people want to tax away a fortune that enables someone to waste their life in this manner, and have the government do something useful with it — but Landsburg argues that it’s practically impossible.

“For the government to consume more goods and services, somebody else must consume fewer,” he writes. But such an heir, he notes, “consumes almost no goods or services whatsoever.” You could give someone else the old Mercedes, but you can’t transform it into health care for a sick child, unless possibly the shock of losing his cars spurs our wastrel heir to retrain as a nurse.

Most people struggle with this because on the individual level, the insight doesn’t hold: If that heir gifted me his trust fund, I could build a big addition with a beautiful walk-in closet, and get a car that wasn’t 16 years old, and maybe a vacation villa in the south of France, and — sorry, where were we?

But at the macroeconomic level, where tax policy operates, there are generally only two ways for some Americans to consume more roads, or health care or what have you: Either we collectively get more efficient at making those things, or someone else must consume less. Which is why Landsburg concludes that any attempt to tax our idle heir “turns into a tax-in-disguise on somebody else” — someone who is already consuming the things we want to redistribute.

(There’s a third possibility: that foreigners could donate stuff to us. I don’t think they will, though.)

Wealth taxes don’t increase productivity; if anything, they discourage capital formation. They do force rich people to reduce their consumption. But since the government can only redistribute or repurpose the stuff that the rich people are themselves consuming, that helps less than the paper wealth might suggest.

Once you’ve got a billionaire’s yacht, and their second or eighth homes, you’ve hit real limits on redistribution, no matter what the accounting entries say. And even an ardent redistributionist might want to take care that the transfers stop with the yacht and don’t inadvertently transform some billionaire entrepreneur’s next start-up into fancy new office furniture for the secretary of housing and urban development.

Which is why the best way to reduce the pernicious effects of unequal wealth might be aggressively taxing consumption.

Consumption taxes are tailored to redistribute real goods and services, and they make investing in the future more attractive, not less. Well-designed ones would probably curtail at least some of the strategies the ultrawealthy currently use to avoid paying taxes. The administrative and economic efficiency of such taxes is one reason that European welfare states tend to raise a lot of revenue from consumption-based, value-added taxes, and comparatively little from wealth taxes.

That said, it would be hard to get there from here. For one thing, the transition would probably be fantastically expensive, since we’d have to make allowances for those who, say, planned their retirement around the old tax code. Consumption taxes also tend to be regressive, though there are ambitious proposals that could theoretically be just as progressive as our current system, while raising at least as much money.

More important, even a highly progressive, revenue-neutral consumption tax would be unpopular in a lot of quarters. It would appall the people who worry that massive fortunes confer massive political influence, even if those fortunes are unspent — and those who hate even the appearance of wealth inequality, even if consumption tells a different story. And it would make new political enemies out of the many Americans who spend every dollar they make.

Yet it remains an idea worth taking seriously, because the United States could use both a better economy and a better tax code, and this is one of the few avenues that promises both.

Read more: